The goal of the current research is to delineate the functional neuroanatomy of stimulus encoding, memory maintenance, and response selection in the human brain using high-field functional MRI. These elemental processes are often subsumed under the heading of working memory-a psychological construct comprising executive functions responsible for the information manipulation and maintenance functions that keep information active over short intervals (Baddeley, 1996). Prior physiological studies in monkeys and neuroimaging studies in humans have implicated dorsolateral prefrontal cortex (d1PFC) the intraparietal sulcus (IPS), and dorsal and ventral posterior brain regions as critical nodes in a neural system supporting working memory. There remains, however, considerable disagreement as to the cognitive processes reflected by activations in these different regions. Our thesis is that working memory tasks do not dissociate important component processes-such as stimulus encoding and maintenance, and response preparation, selection and inhibition. Here we propose two series of experiments to clarify the role of these brain regions in encoding, maintenance, and response processing. In the first experimental series, a modified delayed-response task will be used to segregate memory maintenance of stimulus and response information to successive delay intervals. Of interest is the degree to which the pattern of phasic and sustained delay-interval activity in different cortical regions suggests their differential participation in the encoding and transient memory for domain specific stimulus features and/or response codes. The number of relevant stimulus properties will be systematically manipulated and irrelevant stimulus properties will be introduced to evaluate the sensitivity of regions evincing maintenance-related activity to distraction. In the second series of experiments, continuous choice response tasks will be used to manipulate local stimulus sequence. Behavioral studies have shown that subjects develop expectancies for stimulus (or response) repetitions or alternations based upon the pattern of recent stimulus events, by stimulus probability and inter-stimulus interval, and by other cues that are partially correlated with the stimulus-response mapping. We wish to identify brain regions sensitive to the development of expectancies, and to determine their sensitivity to attention and to task-irrelevant stimulus dimension.